journal citation reports reference guide
TRANSCRIPT
Journal Citation Reports™:Reference Guide
June 2021
2
1. What is the Journal Citation Reports™ (JCR)
2. How is the Journal Impact Factor™ (JIF) calculated3. How is the Journal Citation Indicator calculated4. Timing of Appearance in the JCR5. Title Changes, Mergers and Splits6. JCR and the Web of Science™7. Early Access Policy Change8. Suppressions, Editorial Expressions of Concern9. Forthcoming titles for Reload10.Reload scope and timing11.Acceptable Use
What is the Journal Citation Reports™
3
4Insert footer
How is Journal Citation Reports™ used The Journal Impact Factor™ (JIF) was originally developed by
Drs. Eugene Garfield and Irving H. Sher as a metric to aid in selection of additional journals for the newly created Science Citation Index™.
Today librarians continue to use the Journal Citation Reports™ (JCR) as a tool in building and managing their journal collections. Publishers use the JCR to gauge journal performance and assess their competitors. Researchers use the JCR to identify appropriate journals for publication of their work based on many characteristics.
Starting with the 2021 release, journals from all editions of the Web of Science™ Core Collection will appear in the JCR with journals in the Science Citation Index Expanded™ and Social Sciences Citation Index™ being eligible for JIF metrics.
How is the Journal Impact Factor™calculated
5
6
How is the Journal Impact Factor™ calculated The Journal Impact Factor™ (JIF) is defined as citations to the journal in the
Journal Citation Reports™ (JCR) year to items published in the previous two years, divided by the total number of scholarly items, also known as citable items, (these comprise articles and reviews) published in the journal in the previous two years.
The JCR year is the last complete year within that year’s JCR data set. For example, the JCR year for the 2021 release is 2020.
JIF metrics are available to all journals in the following editions:
• Science Citation Index Expanded™
• Social Science Citation Index™
2021202020192018< 2018
Source paper published in 2020
Cited reference published in 2018 or 2019
7
Journal Impact Factor™Numerator The numerator of the JIF consists of any citation from material
published in the JCR year to material the journal published in the prior two years, regardless of what type of item might be cited. Each cited reference in a scholarly publication is an acknowledgement of influence. JCR therefore aggregates all citations to a given journal in the numerator regardless of cited document type.
The citations that comprise the JIF numerator are drawn from all indexes in the Web of Science Core Collection™: Science Citation Index Expanded™, Social Sciences Citation Index™, Arts & Humanities Citation Index™, Emerging Sources Citation Index™, Conference Proceedings Citation Index, and Book Citation Index™.
2020 JIF =
2020 citations from journals, proceedings, and books to a
journal in 2018 and 2019
Number of articles and reviews (citable items) published in the
journal in 2018 and 2019
Denominator
Numerator
8
Journal Impact Factor™Denominator and Citable Items
Citable items are all items indexed in the Web of Science™ as articles and reviews. Items with any other document type, including editorial material, letter, and meeting abstract, are not included in the denominator.
Document types are assigned at the section level, and each journal is evaluated individually for the characteristics of the items in each section. In our consideration in determining document types for a section, we review the following elements that are typical bibliographic characteristics of scholarly works:
▪ Abstract
▪ Descriptive article titles
▪ Named author with author addresses
No single feature defines a “citable” item –but consideration of all of these, across many examples of the items in each section results in a strongly consistent association between items identified as “citable” (articles and reviews) and items whose citations create the Journal Impact Factor.
2020 JIF =
2020 citations from journals, proceedings, and books to a
journal in 2018 and 2019
Number of articles and reviews (citable items) published in the
journal in 2018 and 2019
Denominator
Numerator
▪ Article length
▪ Cited References
▪ Data content
9
Handling Journal Title Variations in Cited References
There is great variation in how authors format citations, including how cited works (also known as source or journal titles) appear in those citations.
Handling this issue is of critical importance for the JCR to properly attribute and aggregate citations for journal-level metrics. To manage these variants, JCR editors build a series of dictionary entries associating each preferred title with a set of variant forms for the cited work. These entries are derived using direct observation and analysis of cited reference data.
JCR editors have access to more than 1 billion cited references in the Web of Science as well as to specialized citation reports that are prepared for each JCR production cycle. Variants can include alternative abbreviated forms, alternate spellings, common misspellings or typographical errors, part numbers, numbered supplements, and others. Any observed variant that can be associated unambiguously with the covered journal is included in the dictionary.
How is the Journal Citation Indicator Calculated
10
11
How is the Journal Citation Indicator calculated
The Journal Citation indicator is available to all journals in the following editions:
• Science Citation Index Expanded• Social Science Citation Index• Arts & Humanities Citation
Index• Emerging Sources Citation Index
The Journal Citation Indicator, a field-normalized metric, represents the average category-normalized citation impact for papers published in the prior three-year period.
For example, the 2020 Journal Citation Indicator will be calculated for journals that published citable items (i.e., articles or reviews ) in 2017, 2018 and 2019, counting all citations they received from any document indexed between 2017 and 2020.
The value of the Journal Citation Indicator is the mean Category Normalized Citation Impact (CNCI) for all articles and reviews published in the most recent three years (e.g., between 2017 and 2019 for the 2020 indicator value).
Introduction the Journal Citation IndicatorCNCI
More Information:
Timing of Appearance in the Journal Citation Reports™
12
13Insert footer
Timing of Appearance for the Journal ImpactFactor™
Journals only appear in the Journal Citation Reports™ (JCR) when there are three complete and known years of source data; this is always and uniformly the case. This is because the Journal Impact Factor™ (JIF) is defined as citations in the JCR year to items published in the previous two years, divided by the total number of scholarly items published in those previous two years.
New journals, i.e., those starting from volume 1, issue 1, can be included in the JCR after their first year because their prior two years of source data are known—as zero. Journals that have undergone a title change are another instance of this exception. In both scenarios these journals would not receive a JIF but are still listed in the JCR with other metrics to their credit, like the Immediacy Index, that can be calculated on the source data we have for that journal.
Title Changes, Mergers, and Splits
14
Title Changes
15
WoS
Final Year
Select Metrics - No JIF First JIF (Year 1**)
Volume X
Title Change in the Web of Science & JCR
Journal A
WoS
Journal B
Volume X+1 Volume X+2 Volume X+3
JCR No Change
JCR JIF
No Change Select Metrics – JIF*
JIF
Volume X
Year 1 Year 2 Year 3 Year 4
If coverage continues
If any of the criteria occur to a covered journal, it will be considered a new journal.However, coverage will continue under its new title until sufficient material is available for a re-evaluation. If coverage continues upon re-evaluation, the journal will appear in the Web of Science™ and Journal Citation Reports™ (JCR) as noted below.
* A minor journal title change that does not meet the criteria above can continue with the same journal record in the JCR
Criteria*
1. Journal ceases publication and a new journal starts with clear relationship between the ceased and the new journal with new ISSN
2. Journal’s Full Title changes, significantly altering journal title variants and citation unification
3. Journal scope and/or Editorial Oversight change
* Select Metrics – Journal Impact Factor™ (JIF) typically higher** Denominator has Year 1 data only and typically a lower JIF the first year
Journal Mergers
16
WoS
Year 1
Final year
Year 2 Year 3 Year 4
Merged Journals in the Web of Science & JCR
Journal A
JCR No Change No Change Select Metrics – JIF*
Select Metrics - No JIF First JIF (Year 1**)
Volume XWoS
Journal C
Volume X+1 Volume X+2 Volume X+3
JCR JIFJIF
Volume N
WoS
Journal B
JCR No Change No Change Select Metrics - JIF*
Volume N
Final year
If coverage continues
If two or more journals cease publication with a new merged journal starting, the merged journal is considered a new journal. However, coverage will continue under its new title until sufficient material is available for a re-evaluation. If coverage continues upon re-evaluation, the journal will appear in the Web of Science and Journal Citation Reports as noted below.
Criteria1. Two or more journals cease
publication. and2. a new merged journal starts
with clear relationship between the ceased
* Select Metrics – Journal Impact Factor (JIF) typically higher** Denominator has Year 1 data only and typically a lower JIF the first year
17Insert footer
Journal Splits
WoS
Year 1
Final Year
Year 2 Year 3
Select Metrics - No JIF First JIF (Year 1**)
Year 4
Volume X
Journal Splits in the Web of Science & JCR
Journal A
WoS
Journal B
Volume X+1 Volume X+2 Volume X+3
JCR No Change
JCR JIF
No Change Select Metrics – JIF*
JIF
Volume X
Select Metrics - No JIF First JIF (Year 1**)
Volume XWoS Volume X+1 Volume X+2 Volume X+3
JCR JIFJIF
Journal C Year 1 Year 2 Year 3 Year 4
A journal split occurs when a journal ceases publication and splits into two or more new journals. However, coverage will continue under its new title until sufficient material is available for a re-evaluation. If coverage continues upon re-evaluation, the journal will appear in the Web of Science and Journal Citation Reports as noted below.
Criteria1. Existing journal ceases
publication, and2. two or more new journals are
created where there’s a clear relationship between the ceased and new journals, and
3. the new journals have new journal titles and new ISSNs If coverage continues
* Select Metrics – Journal Impact Factor (JIF) typically higher** Denominator has Year 1 data only and typically a lower JIF the first year
Journal Citation Reports™ vs the Web of Science™
18
19Insert footer
Journal Citation Reports™ vs the Web of Science™
A key difference between the Web of Science™ and the Journal Citation Reports™ (JCR) is that the Web of Science is continually updated, while the JCR data represent a snapshot from a specific time, namely, the date when the data were extracted for analysis. This difference reflects the essential nature of each product.
• The Web of Science is an evolving record of the dynamic world of scholarly communication with new source content continually added, including citations.
• JCR, on the other hand, is a report on the citation impact of a defined set of journals at a given moment in time.
Therefore, attempts to use the Web of Science data to replicate JCR metrics will lead to results that can vary from those reported in JCR.
Live Collection vs Snapshot
20Insert footer
Journal Citation Reports™ vs the Web of Science™
If one piece of citation-related data changes in a Web of Science article, only that one article and those to which it connects directly are affected.
If one piece of data changes for the JCR, it changes the overall network’s structure and morphology, sometimes drastically.
• For example, if the item count changes for one journal, not only does that change the Journal Impact Factor for that journal, but also the median and aggregate Journal Impact Factors for any/all categories in which the journal appears. It will also affect the rank in category for many or even all the journals in a category.
It is for these reasons that the JCR is reloaded only once per year, with all data updates occurring simultaneously.
Local Change vs Network of Interconnected Data
Early Access Journal Citation Reports™ Policy Change for 2021
21
22
First, what is Early Access in the Web of Science™…
Compatible Early Access material is journal material that is the Version of Record (VOR), made available early, before it is published in an issue. It has a DOI and distinct early access publication date . When it is later published in an issue, it then has volume, issue, pages, and final publication date data .
2
3
1
*Accepted manuscripts and corrected proofs are not VOR.
1 2
3
23
2021 Early Access Policy Change
Most indexed Early Access items have an early access date and a final publication date in the same calendar year. The new Journal Citation Reports™ (JCR) policy will not change how these items are counted.
For indexed Early Access items where the early access date is in a different calendar year than the final publication date, we will use only the early access date. This will only affect items indexed with an early access date of 2020 or later.
24
Prospective Approach
Based on our analysis, we determined to take a prospective approach, starting with content indexed in 2020.
See Adding Early Access to the Journal Citation Reports for details about this analysis.
Determining JCR year
JCR reviews item dates
Item processed
as Early Access
Early Access Year >=
2020
Use Final Publication Year
Use Final Publication Year*
Use Early Access Year
YesYes
No
No
…
Final Publication
Early Access 2018
Final Publication
2019
Early Access 2020
Final Publication
2021
Paper A Paper B Paper C
✔ ✔
✔
* 2019 and earlier Early Access items with no final publication year are excluded
25
How this will work over time
2020 JIF =
Citations in 2020 to items published in 2018 + 2019
Number of articles & reviews published in 2018 & 2019
2021 JIF =
Citations in 2021 to items published in 2019 + 2020
Number of articles & reviews published in 2019 & 2020
2022 JIF =
Citations in 2022 to items published in 2020 + 2021
Number of articles & reviews published in 2020 & 2021
The years highlighted in blue include EA content as described. By the 2022 Journal Impact Factor™ (JIF), all journals with EA indexed by 2020 will have JIFs reflecting the earliest publication date of their content.
Example
- Citations from the Numerator have a distributed effect and apply to all journals that material cites
- Journals onboarded later will have a similar transition
27
Early Access Citations
Early Access content typically has a higher contribution to JIF when counted by the early access date
Early Access 2020 example with 2021 planned publication
2020 citations to 2019 (7) or 2018 (16)
2021 citations to 2020 (0) or 2019 (7)
Contribution to 2020 JIF: 23
Contribution to 2021 JIF: 7
28
Other FAQs My journal content includes an Early Access date that’s different than the issue. Does this Journal Citation Reports policy apply?
• This policy applies if we are indexing your content when it is Early Access only before it is assigned to an issue. If you publish Early Access but have not been onboarded, the policy will apply in the 2022 release, assuming successful onboarding in 2021.
• It does not apply to Continuous Article Publication (CAP) models or Build Online if the item date and issue date are in the same month/year.
What if my journal(s) has Early Access items from 2019 and earlier that have not been assigned to an issue?
2019 Early Access items and earlier not been assigned to a volume/issue will only contribute to the JCR when the items are indexed in the final volume/issue.
29
Other FAQs What’s the difference between Continuous Article Publication and Build Online for Early Access?
The Early Access indexing model for the Web of Science™ expects to receive a later version with a final publication date, volume, issue and page information from the publisher. In most cases, Continuous Article publications and Build Online publications are published once with the final publication bibliographic data and are incompatible for Early Access indexing. Below are details considered for the Web of Science indexing for these publication models.
– Continuous Article Publication (CAP): Content published through a continuous publication model is comparable to Early Access in that it is published as soon as it is the VOR. Unlike Early Access, it is not later updated with final metadata. Continuous publication material with respect to the Journal Impact Factor (JIF) denominator, including the timing of the cited reference contributions for JIF numerators, are not affected by the Early Access policy change.
– Build Online: This is a publication model whereby journal content is published in a volume and issue while the issue is building to completion. The current Web of Science indexing policy for this model is to index the issue only upon completion. Since Build Online issues typically complete within a calendar year, their published content with respect to the JIF denominator are not affected by the Early Access policy change. There may be exceptions where Build Online journals publish Early Access Content before assignment to an issue. Those are evaluated on a case-by-case basis.
Suppressions and Editorial Expressions of Concern
30
31
What is citation distortion?
Citation distortion is caused by a citation anomaly that exceeds a normal range when compared to a journal’s peers in the same subject category. There are two main types of journal-level citation distortion:
Journal Self-Citation
• A journal has an unusually high percentage of self-citations, as compared to non-self-citations, within the Journal Impact Factor™ (JIF) window
• The distortion in category rank is based on analysis of all journals in all categories of the Journal Citation Reports™ (JCR) ranked both with and without the inclusion of self cites
• The distortion equals the percentage shift in rank with self cites included versus excluded
Citation stacking
• A donor journal cites a recipient journal at an unusually high rate that is concentrated in the Journal Impact Factor window
• Citation stacking involves two or more journals
32
Our principles• Distortion of citations is harmful to the scholarly record
• It creates an inaccurate reflection of the connections between articles and of their contribution to the scholarly network
• Journal self-citation and citation stacking have the potential to distort the citation network, the Web of Science, and the JIF
• The JIF provides an important indicator of a journal’s contribution to scholarly communication and must not be either inadvertently influenced or purposefully gamed
Suppression of Journal Impact Factors
33
Adapting to respond to a changing environment
We updated our methodology and parameters for journal self-citation in 2020
• New, problematic citation networks had surfaced over time
• In prior years' analysis, we focused on extreme outliers. At that time, distortion of the citation network by journal self-citation was both rare and extreme
• The changes make critical adjustments:
• Content: we now include out-going (citing) as well as in-coming (cited) activity as part of the analysis
• Context: We also now make comparisons within each journal’s category/categories, rather than JCR edition, to better account for discipline norms
• Perspective: we now look at the level of distortion of the citation network, not solely at the outcome of that distortion
34
How do we identify journal self-citation distortion?
Suppression decisions are based on analysis of 2020 citation data.
We look at the contribution of self-citations to the following data :
• Total citations (TC)
• Journal Impact Factor
• Rank in category
• % of journal self-citations in in-bound (cited) activity, and % of journal self-citations in out-bound (citing) activity
Suppressed journals represent anomalies in citation behavior and exceed category-based norms.
Journals are suppressed for one year and re-evaluated with the next year’s data.
35
How do we identify citation stacking distortion?
Suppression decisions are based on analysis of 2020 citation data.
We look at the following data:
• Donor as % of Recipient’s Total Citations
• Donor as % of Recipient’s JIF numerator citations
• Concentration of citations exchanged into JIF numerator
• Identification of individual item(s) in Donor Journal with non-scholarly and/or near-exclusive reference to Recipient journal(s)
Suppressed journals represent anomalies in citation behavior and exceed category-based norms.
Both Donor and Recipient journals are suppressed for one year and re-evaluated with the next year’s data.
New journals, where citations are naturally concentrated to recent years, are not suppressed.
36
What do we do when we identify citation distortion?
Journals that show evidence of citation distortion will not receive a Journal Impact Factor.
• We notify publishers of suppressed journals prior to the annual JCR release.
• We report the findings of our analysis without accusations of wrongdoing.
• We give publishers the opportunity to appeal suppression decisions; appeals must be received by the defined deadline and must be data-driven – we do not consider unsupported assertions.
• Suppressed journals continue to be indexed in the Web of Science but may be re-evaluated for continued coverage.
37Insert footer
Journal Citation Reports™Editorial Expressions of Concern
JCR Editorial Expression of Concerns (EEoC) are a signal that we have identified patterns of publication and/or citation that are problematic but for which we do not have a current suppression analysis in place.
• In 2018 and 2019, we identified a set of four journals that exhibited anomalous citation concentration to a fifth, unrelated title. Full investigation resulted in citing titles being suppressed, and/or removed from coverage in the Web of Science in the following years.
• In 2020 and 2021, we identified journals with one or more published items containing a high volume of journal self-citations to the JIF. While not sufficient to trigger journal self-citation suppression, these citations created a visible anomaly in the list of citations contributing to the JIF.
We will continue to use EEoCs, when necessary, to alert users of the need to examine the data contributing to the JIF and JCR metrics.
Forthcoming Titles
38
39Insert footer
Forthcoming Titles• Forthcoming titles are Journal Citation Reports™
(JCR) journals where content has been identified as missing in the final data set where accurate metrics cannot be published in the initial release.
• When available, projected metrics are published in the JCR as a pending correction before the reload update occurs.
• Publishers are notified of titles on the forthcoming list in advance of the June release.
• Missing material not processed by the designated cut-off will not appear in the JCR Reload
Titles not available in the Journal Citation Reports™data year due to missing content
Journal Citation Reports™ Data Reload
40
41Insert footer
Journal Citation Reports™Data Reload
Eligible corrections for Reload are:
– JIF Denominator
• Reclassifying document types for Journal Impact Factor™(JIF) denominator items related to citable vs non-citable counts
• Adding missing items to the denominator count
• Removing duplicated items from the denominator count
• Count adjustment due to publication year correction
– JIF Numerator
• Unifying citations in the original JCR data set for a missing journal title variant, also known as a cited work variant
Corrections out of scope for Reload are as follows:
– JIF Numerator
• Any citation not included in the original JCR data set
• Citation where the cited work is incorrectly attributed to a different journal
A one-time data update for the current Journal Citation Reports data year
Acceptable Use
42
43Insert footer
Acceptable Use Broad re-publication or free distribution of Journal Citation Reports™(JCR) data is not permitted. We do allow distribution of JCR data if it is kept within the subscribing institution. In that case, the JCR should be appropriately acknowledged as the source and the recipients should be made aware of the fact that posting the data to the Web will violate the institution’s license agreement.
Use of the JCR data for marketing, public relations, news stories or publication as part of bibliographic or bibliometric research will require permission from Clarivate. We will request to review the information and data that would be included in the publication and will provide guidance on the citation of JCR as a source.
Permission requests can be submitted through this product support form.
44
Misuse of the Journal Impact Factor™
Responsible use of metrics
The JCR was created as a way to use cited references "as characteristics of the journals they linked" (Garfield, 1975). The Journal Impact Factor™ (JIF) was designed to summarize performance, combining citations across all content to create a generalized, journal-level metric showing a specific type of scholarly use in a single year.
Article citation performance as a part of scholarly assessment is not reflected in the JIF. Metrics that address the output and influence of individual researchers can be analyzed through Web of Science™ or InCites™. Please see this relevant ISI Global Research Report related to this topic, Profiles, not metrics.
Journal coverage and journal reputation
Some websites or journals will falsely claim to be covered in the Web of Science or will reference having an “impact factor”. Misrepresentations of our products and our brands may be an attempt to mislead potential authors into believing that a journal has met our high standards for inclusion in our indexes. Please make use of our publicly available tools to review journals’ claims, like the Web of Science Master Journal List.
Thank you
© 2021 Clarivate. All rights reserved. Republication or redistribution of Clarivate content, including by framing or similar means, is prohibited without the prior written consent of Clarivate. Clarivate and its logo, as well as all other trademarks used herein are trademarks of their respective owners and used under license.